Embodiments are directed to techniques for matching the storage tiers used to back a cloned storage object with the storage tiers used to back the source storage object from which the cloned storage object is to be created. This may be done by determining the storage tier for each source storage extent that backs the source storage object and allocating, for each source storage extent, a corresponding cloned storage extent that matches that source storage extent. Data may then be copied from each storage extent on the source to the corresponding storage extent on the clone.
Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method of cloning a source logical disk backed by a plurality of source storage extents drawn from persistent storage of a data storage apparatus disposed within a single chassis, wherein the plurality of source storage extents that back the source logical disk are drawn from a first storage pool having multiple storage tiers with respective performance levels, the method comprising: accessing metadata of the source logical disk to ascertain, for each of the plurality of source storage extents, a storage tier of the multiple storage tiers of the first storage pool from which that source storage extent is composed; for each of the plurality of source storage extents, allocating a respective cloned storage extent from a second storage pool of the persistent storage of the data storage apparatus to back a cloned logical disk, each of the cloned storage extents being allocated from a storage tier having a performance level that matches the performance level of the storage tier from which the respective source storage extent is composed, wherein the second storage pool has multiple storage tiers with respective performance levels differing from the multiple storage tiers of the first storage pool; and in response to allocating, copying data from the plurality of source storage extents to the respective cloned storage extents, thus providing storage tiering for the cloned logical disk that matches storage tiering for the source logical disk.
This invention relates to data storage systems, specifically methods for cloning logical disks while preserving storage tiering performance characteristics. The problem addressed is the need to maintain consistent performance levels in cloned logical disks when source data is distributed across multiple storage tiers with varying performance levels in a single-chassis storage apparatus. The method involves cloning a source logical disk backed by multiple source storage extents from a first storage pool containing multiple performance-tiered storage tiers. The process begins by analyzing metadata of the source logical disk to determine the specific storage tier of each source storage extent. For each source extent, a corresponding cloned storage extent is allocated from a second storage pool, ensuring the cloned extent is placed in a storage tier with matching performance characteristics to the original. The second storage pool has its own distinct set of performance-tiered storage tiers. After allocation, data is copied from the source extents to their respective cloned extents, resulting in a cloned logical disk with storage tiering that mirrors the original, thereby preserving performance consistency across the cloned system. This approach ensures that performance-critical data remains in high-performance storage tiers while lower-priority data is placed in lower-performance tiers, maintaining the original storage tiering strategy.
2. The method of claim 1 , wherein the source logical disk is realized by a first container file of a first container filesystem, the first container filesystem being backed by storage extents drawn from the first storage pool including the plurality of source storage extents that back the source logical disk; wherein the cloned logical disk is realized by a second container file within a second container filesystem backed by storage extents drawn from the second storage pool including the cloned storage extents that back the cloned logical disk, the first storage pool being managed by a first storage driver and the second storage pool being managed by a second storage driver distinct from the first storage driver; wherein accessing metadata of the source logical disk to ascertain, for each of the plurality of source storage extents, the storage tier from which that source storage extent is composed is performed by the first storage driver; and wherein allocating the respective cloned storage extent for each of the plurality of source storage extents to back the cloned logical disk includes the first storage driver sending an indication of each of the respective ascertained storage tiers to the second storage driver.
This technical summary describes a method for cloning a logical disk across storage pools managed by distinct storage drivers. The invention addresses the challenge of efficiently replicating a logical disk while preserving storage tier information, ensuring performance and cost optimization in heterogeneous storage environments. The method involves a source logical disk implemented as a container file within a first container filesystem, which is backed by storage extents from a first storage pool. This pool includes multiple source storage extents that support the source logical disk. A cloned logical disk is created as a second container file within a second container filesystem, backed by storage extents from a second storage pool. The first and second storage pools are managed by separate storage drivers, ensuring compatibility across different storage management systems. To clone the logical disk, the first storage driver accesses metadata of the source logical disk to determine the storage tier of each source storage extent. This tier information is then communicated to the second storage driver, which allocates corresponding cloned storage extents in the second storage pool. The cloned storage extents are selected based on the same storage tiers as the source extents, maintaining performance characteristics and cost efficiency. This approach enables seamless cloning across diverse storage infrastructures while preserving storage tiering policies.
3. The method of claim 2 wherein copying data from the plurality of source storage extents to the respective cloned storage extents includes the first storage driver writing the data from each source storage extent to a shared cache and the second storage driver copying the data that was written to the shared cache from the shared cache to the respective cloned storage extents.
This invention relates to data cloning in storage systems, specifically improving efficiency in copying data from source storage extents to cloned storage extents. The problem addressed is the performance overhead and resource contention that occurs when multiple storage drivers independently access source data during cloning operations, leading to inefficiencies and potential bottlenecks. The solution involves a method where a first storage driver writes data from each source storage extent to a shared cache, and a second storage driver then copies the data from the shared cache to the respective cloned storage extents. This approach centralizes data access through the shared cache, reducing redundant reads from the source storage and minimizing contention between storage drivers. The shared cache acts as an intermediary, allowing the second storage driver to retrieve data more efficiently without directly accessing the source extents. This method ensures that data is transferred in a coordinated manner, improving overall cloning performance and reducing system resource usage. The technique is particularly useful in environments where multiple storage drivers operate concurrently, such as in virtualized or distributed storage systems.
4. The method of claim 2 , wherein the metadata of the source logical disk is stored on and accessed from a metadata portion of the first container filesystem, the metadata portion of the first container filesystem being stored on the persistent storage of the data storage apparatus; and wherein accessing metadata of the source logical disk includes accessing a tiering table stored within the metadata portion of the first container filesystem stored on the persistent storage of the data storage apparatus, the tiering table including a set of mappings, each mapping of the set corresponding to a respective source storage extent of the plurality of source storage extents and pointing to the storage tier from which that source storage extent is composed.
The invention relates to data storage systems, specifically methods for managing and accessing metadata in a container filesystem environment. The problem addressed involves efficiently storing and retrieving metadata for logical disks within a container filesystem, particularly in systems where data is distributed across multiple storage tiers. Traditional approaches may suffer from inefficiencies in metadata access, leading to performance bottlenecks or complex management overhead. The method involves storing metadata for a source logical disk within a metadata portion of a container filesystem, which is itself stored on persistent storage of a data storage apparatus. This metadata includes a tiering table that maps source storage extents to their respective storage tiers. Each entry in the tiering table corresponds to a specific source storage extent and indicates the storage tier from which that extent is composed. By maintaining this tiering table within the container filesystem's metadata portion, the system ensures that metadata access is streamlined and that the relationships between storage extents and their tiers are easily retrievable. This approach simplifies data management and improves performance by reducing the need for external metadata lookups or complex tiering logic. The method is particularly useful in environments where data is dynamically moved between different storage tiers based on performance or cost considerations.
5. The method of claim 1 wherein allocating the respective cloned storage extent for each of the plurality of source storage extents to back the cloned logical disk includes allocating each cloned storage extent of a closest available type as the ascertained storage tier of the respective source storage extent.
This invention relates to data storage systems, specifically methods for efficiently allocating storage resources in a cloning operation. The problem addressed is the inefficient use of storage tiers when creating cloned logical disks from source storage extents, which can lead to suboptimal performance and resource utilization. The method involves cloning a logical disk by allocating storage extents from a storage system to back the cloned logical disk. Each source storage extent is analyzed to determine its storage tier, which refers to the performance characteristics of the underlying storage media (e.g., SSD, HDD, or archival storage). When allocating cloned storage extents to back the cloned logical disk, the system ensures that each cloned extent matches the storage tier of its corresponding source extent as closely as possible. If an exact match is unavailable, the system selects the closest available storage tier to maintain performance consistency. The method also includes dynamically adjusting the allocation of cloned storage extents if the storage system's configuration changes, such as when new storage tiers become available or existing ones are modified. This ensures that the cloned logical disk continues to operate efficiently over time. The approach optimizes storage resource usage by aligning the performance characteristics of the cloned extents with those of the source extents, improving overall system efficiency and reducing unnecessary data movement.
6. The method of claim 1 wherein the multiple storage tiers includes a first storage tier made up of solid state storage drives and a second storage tier made up of magnetic disk drives.
This invention relates to a multi-tiered storage system designed to optimize data storage efficiency and performance. The system addresses the challenge of balancing cost and speed in data storage by utilizing different types of storage media for different data access needs. The primary storage tier consists of solid-state storage drives, which provide high-speed access for frequently used or critical data. The secondary storage tier comprises magnetic disk drives, which offer lower-cost, high-capacity storage for less frequently accessed data. The system dynamically manages data placement between these tiers based on access patterns, ensuring that high-priority data resides in the faster solid-state drives while less critical data is stored in the slower but more economical magnetic disk drives. This tiered approach improves overall system performance by reducing latency for frequently accessed data while minimizing storage costs. The invention also includes mechanisms to monitor data access patterns and automatically migrate data between tiers as needed, ensuring optimal performance without manual intervention. The combination of solid-state and magnetic storage allows the system to adapt to varying workload demands efficiently.
7. The method of claim 1 wherein the plurality of source storage extents and the respective cloned storage extents all share a fixed extent size within a range of between 256 megabytes and 1 gigabyte.
This invention relates to data storage systems, specifically methods for managing storage extents in a cloning operation. The problem addressed is the inefficiency in storage allocation and performance when cloning data, particularly when dealing with varying extent sizes that can lead to fragmented storage or suboptimal performance. The method involves cloning data from a plurality of source storage extents to respective cloned storage extents, where all extents—both source and cloned—share a fixed extent size. This fixed size is constrained within a specific range, between 256 megabytes and 1 gigabyte. By enforcing a uniform extent size, the system ensures consistent storage allocation, reduces fragmentation, and improves performance during cloning operations. The fixed size range balances the need for granularity in storage management with the efficiency of handling larger blocks of data. The method may also include dynamically adjusting the extent size within the specified range based on system conditions, such as available storage capacity or performance requirements. This flexibility allows the system to optimize storage usage while maintaining the benefits of fixed-size extents. The approach is particularly useful in environments where large-scale data cloning is required, such as in virtualization, backup systems, or cloud storage solutions.
8. The method of claim 1 wherein the method further comprises, after copying the data from the plurality of source storage extents to the respective cloned storage extents, refraining from keeping the plurality of source storage extents synchronized with the respective cloned storage extents.
This invention relates to data storage systems, specifically methods for managing cloned storage extents. The problem addressed is the unnecessary overhead and complexity of maintaining synchronization between source storage extents and their cloned copies when such synchronization is not required. The method involves copying data from multiple source storage extents to respective cloned storage extents. After this copying process, the system intentionally avoids keeping the source and cloned extents synchronized. This means that any subsequent changes to the source extents are not automatically reflected in the cloned extents, and vice versa. By refraining from synchronization, the system reduces resource consumption and simplifies management, particularly in scenarios where the cloned extents are intended for independent use, such as backups, snapshots, or test environments. The method ensures that the cloned extents are initially identical to the source extents but then operate independently. This approach is useful in storage systems where maintaining synchronization would be unnecessary or undesirable, such as when the cloned extents are used for archival purposes or when the source data is expected to change frequently. The technique optimizes performance and storage efficiency by eliminating the need for ongoing synchronization operations.
9. A computer program product comprising a non-transitory computer-readable storage medium storing a set of instructions, which, when performed by computing circuitry of a data storage apparatus disposed within a single chassis, cause the data storage apparatus to clone a source logical disk backed by a plurality of source storage extents drawn from persistent storage local to the data storage apparatus, the plurality of source storage extents that back the source logical disk being drawn from a first storage pool having multiple storage tiers with respective performance levels, by: accessing metadata of the source logical disk to ascertain, for each of the plurality of source storage extents, a storage tier of the multiple storage tiers of the first storage pool from which that source storage extent is composed; for each of the plurality of source storage extents, allocating a respective cloned storage extent from a second storage pool of the persistent storage of the data storage apparatus to back a cloned logical disk, each of the cloned storage extents being allocated from a storage tier having a performance level that matches the performance level of the storage tier from which the respective source storage extent is composed, wherein the second storage pool has multiple storage tiers with respective performance levels differing from the multiple storage tiers of the first storage pool; and in response to allocating, copying data from the plurality of source storage extents to the respective cloned storage extents, thus providing storage tiering for the cloned logical disk that matches storage tiering for the source logical disk.
This invention relates to data storage systems and addresses the challenge of efficiently cloning logical disks while preserving performance characteristics. The system involves a data storage apparatus within a single chassis that clones a source logical disk backed by multiple source storage extents from a first storage pool. The first storage pool comprises multiple storage tiers with varying performance levels. The cloning process begins by accessing metadata of the source logical disk to determine the storage tier of each source storage extent. For each source extent, a corresponding cloned storage extent is allocated from a second storage pool, which also has multiple storage tiers with different performance levels. The cloned extents are allocated from tiers that match the performance levels of the original source extents. After allocation, data is copied from the source extents to the cloned extents, ensuring that the cloned logical disk maintains the same storage tiering and performance characteristics as the source logical disk. This approach enables efficient cloning while preserving performance consistency across different storage pools.
10. The computer program product of claim 9 wherein allocating the respective cloned storage extent for each of the plurality of source storage extents to back the cloned logical disk includes allocating each cloned storage extent of a closest available type as the ascertained storage tier of the respective source storage extent.
This invention relates to data storage systems, specifically optimizing storage tier allocation for cloned logical disks. The problem addressed is inefficient storage resource utilization when creating clones of source storage extents, where the cloned data may not be optimally placed in the most appropriate storage tier based on performance or cost requirements. The system involves a method for creating cloned logical disks from source storage extents, where each source extent resides in a specific storage tier (e.g., high-performance SSD, cost-effective HDD, or cloud storage). When cloning, the system allocates a cloned storage extent for each source extent, ensuring the cloned extent matches the storage tier of the original. If the exact tier is unavailable, the system selects the closest available tier to maintain performance and cost efficiency. This ensures that cloned data retains similar performance characteristics to the original while minimizing unnecessary resource consumption. The process includes identifying the storage tier of each source extent, determining available storage tiers for cloning, and allocating the closest matching tier for each cloned extent. This approach prevents performance degradation or excessive costs by avoiding arbitrary tier assignments. The system dynamically adjusts allocations based on real-time storage availability, ensuring optimal resource usage. This method is particularly useful in virtualized environments where storage cloning is frequent, such as in backup, disaster recovery, or development testing scenarios.
11. The computer program product of claim 9 wherein the multiple storage tiers includes a first storage tier made up of solid state storage drives and a second storage tier made up of magnetic disk drives.
A multi-tiered storage system is designed to optimize data storage efficiency and performance by utilizing different types of storage media. The system includes at least two distinct storage tiers: a first tier composed of solid-state storage drives, which provide high-speed access and low latency, and a second tier composed of magnetic disk drives, which offer higher capacity at a lower cost. The system dynamically allocates data across these tiers based on factors such as access frequency, performance requirements, and cost considerations. Solid-state drives are used for frequently accessed or performance-critical data, while magnetic disk drives handle less frequently accessed or bulk storage needs. This tiered approach improves overall system efficiency by balancing cost, performance, and capacity requirements. The system may also include additional storage tiers or hybrid configurations to further enhance flexibility and scalability. The invention is particularly useful in data centers, cloud storage, and enterprise environments where optimizing storage resources is critical.
12. An apparatus comprising: a chassis; persistent storage disposed within the chassis; and processing circuitry coupled to memory disposed within the chassis configured to clone a source logical disk backed by a plurality of source storage extents drawn from the persistent storage, the plurality of source storage extents that back the source logical disk being drawn from a first storage pool having multiple storage tiers with respective performance levels, by: accessing metadata of the source logical disk to ascertain, for each of the plurality of source storage extents, a storage tier of the multiple storage tiers of the first storage pool from which that source storage extent is composed; for each of the plurality of source storage extents, allocating a respective cloned storage extent from a second storage pool of the persistent storage to back a cloned logical disk, each of the cloned storage extents being allocated from a storage tier having a performance level that matches the performance level of the storage tier from which the respective source storage extent is composed, wherein the second storage pool has multiple storage tiers with respective performance levels differing from the multiple storage tiers of the first storage pool; and in response to allocating, copying data from the plurality of source storage extents to the respective cloned storage extents, thus providing storage tiering for the cloned logical disk that matches storage tiering for the source logical disk.
This invention relates to data storage systems and specifically addresses the challenge of maintaining performance characteristics when cloning logical disks across different storage pools with varying performance tiers. The apparatus includes a chassis housing persistent storage and processing circuitry coupled to memory. The system clones a source logical disk, which is backed by multiple source storage extents drawn from a first storage pool. This first pool comprises multiple storage tiers, each with distinct performance levels. The cloning process involves accessing metadata of the source logical disk to determine the storage tier of each source storage extent. For each extent, the system allocates a corresponding cloned storage extent from a second storage pool, ensuring that the performance level of the cloned extent matches that of the source extent. The second storage pool also has multiple tiers, but these differ in configuration from the first pool. After allocation, data is copied from the source extents to the cloned extents, preserving the original storage tiering structure. This approach ensures that the cloned logical disk maintains the same performance characteristics as the source disk, even when the underlying storage infrastructure differs. The solution is particularly useful in environments where storage resources are dynamically allocated or where data migration between pools is required without degrading performance.
13. The apparatus of claim 12 wherein allocating the respective cloned storage extent for each of the plurality of source storage extents to back the cloned logical disk includes allocating the each cloned storage extent of a closest available type as the ascertained storage tier of the respective source storage extent.
A system and method for managing storage resources in a data storage environment involves allocating cloned storage extents to back cloned logical disks. The system identifies a plurality of source storage extents, each associated with a specific storage tier based on performance characteristics such as speed or cost. When creating a cloned logical disk, the system allocates a cloned storage extent for each source storage extent, ensuring that the cloned extent matches the storage tier of the corresponding source extent. If the exact storage tier is unavailable, the system selects the closest available type to maintain performance consistency. This approach optimizes storage resource allocation by dynamically assigning storage extents based on tier availability while preserving the performance characteristics of the original data. The system may also include mechanisms for tracking storage tier assignments and adjusting allocations as storage resources change. The solution addresses the challenge of efficiently managing cloned storage in heterogeneous storage environments, ensuring that performance expectations are met while maximizing resource utilization.
14. The apparatus of claim 12 wherein the multiple storage tiers includes a first storage tier made up of solid state storage drives of the persistent storage and a second storage tier made up of magnetic disk drives of the persistent storage.
A multi-tiered storage system is designed to optimize data storage efficiency and performance by utilizing different types of storage media. The system includes at least two distinct storage tiers: a first tier composed of solid-state storage drives (SSDs) and a second tier composed of magnetic disk drives (HDDs). The SSDs provide high-speed, low-latency access for frequently accessed or performance-critical data, while the HDDs offer higher capacity and lower cost for less frequently accessed data. The system dynamically manages data placement across these tiers based on access patterns, ensuring that frequently accessed data resides in the faster SSD tier while less critical data is stored in the HDD tier. This tiered approach improves overall system performance by reducing latency for critical operations while maintaining cost efficiency through the use of lower-cost HDDs for bulk storage. The system may also include additional storage tiers or hybrid configurations to further optimize performance and cost. The storage tiers are part of a persistent storage system, ensuring data durability and reliability across all storage media types.
Cooperative Patent Classification codes for this invention. Click any code to explore related patents in that topic.
September 30, 2015
December 3, 2019
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.